To become a data analyst, learn Excel, SQL and one visualisation tool such as Power BI or Tableau, add basic statistics and some Python, and build three to four projects that answer real business questions. Data analytics is one of the most accessible ways into IT for graduates from any stream, because it rewards clear thinking and communication as much as technical skill. This roadmap shows what to learn, in what order, and how to prove your skills to employers.
What does a data analyst do?
A data analyst turns raw data into answers that help people make decisions. Typical questions include:
- Why did sales drop in the south region last quarter?
- Which marketing channel brings the customers who stay longest?
- How long does it take to resolve a support ticket, and where are the delays?
The work involves collecting data, cleaning it, analysing it, building dashboards and explaining findings to non-technical colleagues.
The data analyst skill stack
| Skill | Why it matters | Time to learn basics |
|---|---|---|
| Excel | Still used in almost every company for quick analysis | 2 to 3 weeks |
| SQL | Pulls data from company databases; tested in most interviews | 4 to 5 weeks |
| Power BI or Tableau | Builds interactive dashboards for managers | 3 to 4 weeks |
| Statistics | Helps you avoid wrong conclusions | 3 weeks |
| Python (pandas) | Handles larger data and automates repeated work | 5 to 6 weeks |
| Communication | An insight nobody understands has no value | Practise throughout |
The data analyst roadmap
Step 1: Excel (weeks 1 to 3)
- Clean data: remove duplicates, text functions, find and replace, data validation.
- Formulas:
IF,SUMIFS,COUNTIFS,XLOOKUP(orVLOOKUP),INDEX/MATCH. - Pivot tables and pivot charts.
- Basic charts and conditional formatting.
Step 2: SQL (weeks 4 to 8)
SQL is the single most important technical skill for analysts. Learn to filter, aggregate and join tables, then move on to subqueries, CTEs and window functions.
-- Top 5 products by revenue last month
SELECT p.product_name,
SUM(o.quantity * o.unit_price) AS revenue
FROM orders o
JOIN products p ON p.product_id = o.product_id
WHERE o.order_date >= DATE '2026-10-01'
AND o.order_date < DATE '2026-11-01'
GROUP BY p.product_name
ORDER BY revenue DESC
LIMIT 5;Practise on free sample databases and write at least 100 queries before you start interviewing.
Step 3: Power BI or Tableau (weeks 9 to 12)
Pick one. Power BI is very common in companies that use Microsoft 365; Tableau is popular in analytics-heavy teams. Learn to connect data sources, model relationships, write basic measures (DAX in Power BI), and design a clean dashboard with filters.
A good dashboard answers a few clear questions. Put the most important number at the top left, use consistent colours, and remove anything that does not help a decision.
Step 4: Statistics (weeks 13 to 15)
- Averages, spread, percentiles and outliers.
- Distributions and sampling.
- Correlation vs causation.
- A/B testing and statistical significance, explained in plain language.
Step 5: Python for analysis (weeks 16 to 21)
Once Excel becomes slow or repetitive, Python helps. Learn the basics from our Python plan, then pandas for cleaning and grouping data, and matplotlib or Seaborn for charts.
import pandas as pd
df = pd.read_csv("support_tickets.csv", parse_dates=["opened", "closed"])
df["hours_to_close"] = (df["closed"] - df["opened"]).dt.total_seconds() / 3600
print(df.groupby("category")["hours_to_close"].median().sort_values())Portfolio projects employers notice
- Sales performance dashboard in Power BI or Tableau with region, product and month filters, plus three written insights.
- SQL case study: answer ten business questions on a public dataset and publish the queries with explanations.
- Customer analysis in Python: segment customers by purchase behaviour and recommend actions for each segment.
- An end-to-end project: data from a public source, cleaned in Python, stored in a database, analysed with SQL and shown in a dashboard.
Write each project up like a short business report: the question, the approach, the findings and a recommendation.
Data analyst vs data scientist
Analysts mostly explain what happened and why; data scientists build models that predict what will happen. Analytics is a great first job, and many analysts later move into data science by adding machine learning skills.
Ready for interviews?
Once you are comfortable with the topics above, practise with our SQL interview questions for data analysts, with clear answers you can explain in your own words.
Frequently asked questions
Can a commerce or arts graduate become a data analyst?
Yes. Business understanding is a real advantage in analytics. Learn Excel, SQL and a dashboard tool, and build projects in a domain you understand, such as finance or retail.
Is coding required for data analytics?
SQL is essential. Python is very useful and increasingly expected for mid-level roles, but many entry-level roles focus on Excel, SQL and Power BI.
Power BI or Tableau: which should I learn?
Check job listings in your city. In many Indian companies, Power BI appears more often because of Microsoft 365 adoption. The concepts transfer, so learning the second tool later is quick.
How long does it take to become a data analyst?
Around four to six months of steady practice is realistic for an entry-level role, if you build projects along the way.
Keep learning with our data analytics tutorials, and see how to present your projects in our career guides.
